Antisense mrna-mediated Bacteriophage Resistance in Lactococcus lactis subsp. lactis

Transcription

1 APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Apr. 1991, p /91/ $02.00/0 Copyright 1991, American Society for Microbiology Vol. 57, No. 4 Antisense mrna-mediated Bacteriophage Resistance in Lactococcus lactis subsp. lactis SUNG GUK KIM AND CARL A. BATT* Department of Food Science, Cornell University, Ithaca, New York Received 31 October 1990/Accepted 16 January 1991 Resistance to a broad class of isometric bacteriophages that infect strains of Lactococcus lactis has been engineered into a dairy starter by expression of antisense mrna targeted against a conserved bacteriophage gene. Maximum protection is obtained only when the entire 1,654-bp coding sequence for a 51-kDa protein is positioned in the antisense orientation with respect to a promoter sequence that functions in L. lactis subsp. lactis. Expression of the antisense mrna results in more than 99% reduction of the total number of PFU. Plaques that do form are characterized by their relatively small size and irregular shape. A variety of truncated genes, including the open reading frame expressed *n the sense orientation, fail to provide any significant measure of resistance as compared with that of the intact open reading frame. Southern hybridization with probes specific for the conserved region reveal that the antisen e plasmid constructs are maintained despite the presence of a large complement of other indigenous plasmids. Strains harboring the antisense mrna plasmid construct grow and produce acid at a rate equivalent to that of the host strain alone, suggesting that antisense expression is not deleterious to normal cellular metabolism. Bacteriophages exist that can and will infect lactic acid bacteria, resulting in lysis or near total lysis of the culture (8). Lactic acid bacteria as either a deliberate or environmental inoculum are used for the production of a number of fermented products consumed by humans and animals. Since the realization that bacteriophages can have a detrimental effect on dairy fermentations, a number of different approaches have been employed to reduce their proliferation (23). These include the implementation of closed vat systems, improved hygienic practices, starter culture rotation, and the formulation of a starter medium that suppresses the ability of bacteriophages to replicate. Bacteriophage-resistant starter cultures have also been selected by direct screening for mutants that can withstand bacteriophage challenge (13, 20). Host-mediated bacteriophage resistance appears to involve either reduced adsoqption, restriction/modification, or bacteriophage maturation. Frequently the development of resistance is mediated by the formation of a cointegrate structure between two plasmids upon conjugation (10). Antisense mrna results from the transcription of the DNA strand complementary to the strand that codes for the protein. Although antisense mrna has been reported to block the expression of a targeted gene in a number of systems, the mechanism is not completely understood. Antisense mrna has been employed to investigate the expression of bacterial genes (16), mammalian genes (7), and, more recently, plant genes (4). It may also regulate the expression of certain genes, for example, the Escherichia coli ompf gene (15). The use of antisense mrna to block bacteriophage or viral infection has been reported for the Rous sarcoma virus (17) and the E. coli SP bacteriophage (3). In this latter study, up to 96% inhibition of PFU was observed. Finally, antisense mrna has been demonstrated to inhibit the ability of the E. coli bacteriophage k to form lysogens by expressing the oop RNA (18). We have expressed the * Corresponding author antisense strand of a 1,654-bp fragment previously found to reside in a number of Lactococcus lactis subsp. lactis bacteriophages. The expression of this antisense mrna appears to block effectively the infection of a number cf related bacteriophages. MATERIALS AND METHODS Brcterial strains and media. L. lactis subsp. lactis starter cult ire CC9 was obtained from a commercial starter inoculun k ndly provided by D. Barbano (Cornell University). All L. lactis subsp. lactis bacteriophages were isolated from cheese wheys obtained from two commercial plants in New York State (9). E. coli JM105 (Alac pro thi stra enda sbcb15 hsdr4 [F' trad36 proab+ laciq ZM15]) was used as a host for initial cloning and sequence analysis. The vectors used were puc12/13, M13mpl8/19 (22), pgkv259 (from J. Kok [21]), and pbluescript (Stratagene, La Jolla, Calif.). LB medium was used for the propagation of E. coli from which plasmid DNA was purified (12). M17 medium with 0.5% glucose (M17G) was used for propagation of L. lactis subsp. lactis strains (19). Sterilized 10% nonfat dry milk was used to grow L. lactis subsp. lactis or acidification trials. Where noted, agar (15 g/liter), ampicillin (50,ug/ml), 5-bromo-4-chloro-3-indolyl-p-D-galactoside (40,ug/ml), and isopropyl-p-d-thiogalactopyranoside (2 mm) were added. Bacteriophage isolation and quantitation. Bacteriophages were isolated from cheese whey collected over a period of 6 years (1983 through 1989). The numbers of PFU were quantified by standard plaque assays employing an overlay culture of L. lactis subsp. lactis in M17 medium with 0.7% agar. Plates were incubated for 24 h at 30 C, and the resulting plaques were counted. All bacteriophages were purified by three successive plaque isolations. Bacteriophage stocks were prepared by polyethylene glycol sodium chloride precipitation as described by Yamamoto and Alberts (24). Bacteriophages were purified further with cesium chloride density gradients as described for bacteriophage K (12).

2 1110 KIM AND BATT DNA isolation and manipulations. Plasmid DNA was isolated from E. coli by the boiling method for small-scale preparation and by a modification of the alkaline lysis protocol for large-scale cesium chloride-ethidium bromide density gradient preparations (12). Plasmid DNA was isolated from L. lactis subsp. lactis by the method of Anderson and McKay (2). DNA was isolated from L. lactis subsp. lactis bacteriophages as described by Maniatis et al. (12) for E. coli bacteriophage X. All restriction and other modifying enzymes were purchased from New England BioLabs (Beverly, Mass.), Amersham Corp. (Arlington Heights, Ill.), and Boehringer-Mannheim (Indianapolis, Ind.) and used according to the instructions of the manufacturers. Southern hybridization was performed as described by Maniatis et al. (12) with GeneScreen Plus membranes. The probe was a synthetic 17-mer specific for nucleotides (nt) 60 through 44 (noncoding strand 5'-CGCCATTGTCATCTCAC-3') of the 1,654-bp fragment (9). Bacterial transformations. E. coli JM105 was transformed by the method of Hanahan (5). L. lactis subsp. lactis CC9 was transformed by the method of Kok et al. (11), and transformants were selected on SM17G (25% sucrose added as an osmotic stabilizer) agar with erythromycin (4,ug/ml). RESULTS Construction of vectors for expression of antisense mrna. Previously, we identified a genetic element that is conserved in the genomes of numerous L. lactis subsp. lactis bacteriophage isolates, and we determined its nucleotide sequence (9). Approximately 95 to 99% of all L. lactis subsp. lactis bacteriophages collected between 1983 and 1989 from two geographically distinct sources carry this conserved DNA fragment. This 1.6-kb EcoRI fragment has a single large 1,356-bp open reading frame coding for a 51-kDa protein. The enc6ded protein is highly charged and shares some homology with yeast translation initiation factor. In addition, there is a potential zinc binding domain within this protein, similar to those observed in genes from bacteriophages T4 and T7, although the exact function of this protein is not known. A number of lactococcal promoter sequences have been isolated by the use of promoter screening vectors, including pgkv210. A L. lactis subsp. cremoris Wg2 proteinase promoter has been characterized that, when fused to the cat-86 gene, confers chloramphenicol resistance (14). The vector pgkv259 was used to express several fragments derived from the conserved 1.6-kb EcoRI fragment, including the following: (i) the entire fragment in both the sense and antisense orientations, (ii) a 695-bp EcoRI-EcoRV fragment in the antisense orientation, and (iii) a 422-bp EcoRV- ClaI fragment in the antisense orientation (Fig. 1). Each of these fragments was cloned separately into pgkv259 by transformation of E. coli JM105, and the expected restriction patterns were confirmed. Recombinant plasmids were isolated from E. coli JM105 and used to transform L. lactis subsp. lactis CC9 with selection for Emr. Effect of antisense mrna on bacteriophage resistance. Each of the individual constructs was tested for its ability to confer resistance to bacteriophage infection by quantifying the number of PFU observed on the strain carrying psgk1.6r as compared with that on either the strain alone or the strain carrying only pgkv259 (Table 1). When challenged with 108 PFU of bacteriophage (mi7-9, an efficiency of plating equivalent to that of the vector-free control was observed in all L. lactis subsp. lactis strains containing psgk1.6r psgk1.6 psgk1.or psgko.8r APPL. ENVIRON. MICROBIOL. EcoRI EcoRV Hindil Cial EcQRI 1 Kb FIG. 1. Restriction map of antisense mrna constructs. Arrows denote the direction of transcription relative to that of the promoter in pgkv259. any antisense construct, with the exception of psgk1.6r. This latter construct contains the 1.6-kb fragment from L. lactis subsp. lactis bacteriophage (mi7-9, in the antisense orientation. When L. lactis subsp. lactis bacteriophage (mi7-9 was used to infect L. lactis subsp. lactis CC9(pSGK1.6R), an efficiency of plating of approximately 0.4% was observed. The plaques that did form were characteristically small (less than 0.2 to 0.5 mm in diameter) and irregularly shaped. In contrast, strains containing any other construct exhibited 2- to 3-mm plaques that were indistinguishable from the plaques produced on the parental strain. Bacteriophage isolated from these petite plaques produced by L. lactis subsp. lactis CC9(pSGK1.6R) formed normal plaques when subsequently tested against the vector-free strain. This suggests that the presence of psgk1.6r did not select for mutants. A number of other bacteriophage isolates, including 4mi7-1 and imi3-1, which contain the 1.6-kb EcoRI fragment, were tested for their, ability to form plaques on L. lactis subsp. lactis CC9(pSGK1.6R). All exhibited a 99.0 to 99.8% reduction in PFU. In contrast, bacteriophage 4C-8, which, on the basis of Southern hybridization with a 32p labeled 1.6-kb EcoRI fragment as a probe, does not appear to contain this conserved region (9), was able to infect all strains, including L. lactis subsp. lactis(psgk1.6r). It would appear, therefore, that psgk1.6r confers resistance only against bacteriophages that carry the gene coding for the 51-kDa protein. The effect of psgk1.6r on the growth kinetics of L. lactis subsp. lactis CC9 and its ability to inhibit bacteriophage infection in broth culture was also measured (Fig. 2). 4mi7-9-infected cultures of L. lactis subsp. lactis CC9 alone and carrying pgkv259 and psgk1.6r were grown in M17G TABLE 1. Effect of antisense mrna construct psgk1.6r on the ability of various bacteriophage isolates to form plaques on L. lactis subsp. lactis CC9 Bacteriophage No. of PFU formed Host alone pgkv259 psgk1.6r (mi7-9 1 x x x 105 (99.6)a 4mi7-1 2 x 109 NDb 2 x 107 (99.0) (mi3-1 2 x 108 ND 5 X 105 (99.8) (C-8 1 x x x 109 (0) a Numbers within parentheses indicate the percent inhibition of L. lactis subsp. lactis CC9(pSGKl.6R) compared with that of the host strain L. lactis subsp. lactis CC9 alone. b ND, Not determined. I

3 VOL. 57, (o Time (min) FIG. 2. Growth kinetics of L. lactis subsp. lactis CC9 infected with (mi7-9 at a multiplicity of infection of 0.01 in M17G broth at 30 C. Symbols: *, L. lactis subsp. lactis CC9 alone, uninfected; *, L. lactis subsp. lactis CC9 alone, infected; A, L. lactis subsp. lactis CC9(pGKV259), infected; 0, L. lactis subsp. lactis CC9 (psgk1.6r), infected. broth, and the optical density was followed with time. Lysis was observed approximately 120 min after infection in both the host strain alone and the strain carrying pgkv259. In comparison, L. lactis subsp. lactis CC9(pSGK1.6R) infected at a multiplicity of infection of 0.01 grew at a rate similar to that of the uninfected host strain. No significant lysis of the infected L. lactis subsp. lactis CC9(pSGK1.6R) culture was observed after 30 h of incubation, supporting the previous observation of the inhibitory effect of the antisense mrna construct. The classical measure of the activity of a starter culture is its ability to acidify milk. To test the effect of psgk1.6r on the ability of L. lactis subsp. lactis CC9 to acidify milk, cultures were inoculated into milk, and the change in titratable acidity was measured with time (Fig. 3). The presence of psgk1.6r did not significantly inhibit the production of acid, nor did the addition of Xmi7-9 adversely affect it. In contrast, the plasmid-free strain did not produce any acid when infected with Xmi7-9 because of the rapid lysis of the culture. Maintenance of psgk1.6r in L. lactis subsp. lactis CC9 transformants. L. lactis subsp. lactis CC9, like other strains of L. lactis subsp. lactis, harbors approximately 7 to 10 different plasmids (14). Therefore, direct plasmid analysis of L. lactis subsp. lactis CC9 transformants could not be used to determine the integrity of psgk1.6r due to the complexity of the plasmid profile. However, it is clear that the plasmid profiles of the parental strain and its transformants are similar, and no gross rearrangements or deletions are obvious (data not shown). Southern hybridization of the total plasmid contents of L. lactis subsp. lactis CC9 digested with EcoRI was performed with a 32P-labeled synthetic 17-mer probe (Fig. 4). A distinct 1.6-kb EcoRI fragment was observed in all L. lactis subsp. lactis CC9 transformants carrying psgk1.6r. As expected, no homologous sequences were observed in either the control parental strain or the strain carrying the vector pgkv S co a) 0.50 ca ANTISENSE mrna Time (hrs) FIG. 3. Rate of acidification of 10% nonfat dry milk by L. lactis subsp. lactis CC9 alone and with psgk1.6r. Nonfat dry milk was inoculated at 2.5% with an overnight culture of L. lactis subsp. lactis CC9 alone or carrying psgk1.6r. Titratable acidity is presented as the percent lactic acid. Symbols: -, L. lactis subsp. lactis CC9 infected with (mi7-9 at a multiplicity of 0.001; - -, L. lactis subsp. lactis CC9, uninfected; - -, L. lactis subsp. lactis (psgk1.6r) infected with (mi7-9 at a multiplicity of 0.001;..., L. lactis subsp. Iactis(pSGK1.6R), uninfected. DISCUSSION Resistance to bacteriophage infection can develop in a previously sensitive host due to a mutational event that effects adsorption, replication, or maturation of the invading bacteriophage. In contrast, new or modified restriction systems can be acquired that cleave bacteriophage DNA upon entry into the cell. All of these mechanisms can develop naturally or be introduced in a purposeful manner. Unfortunately, bacteriophages can also evolve to circumvent these FIG. 4. Southern hybridization of L. lactis subsp. lactis transformants carrying psgk1.6r. Total plasmid DNA was isolated and restricted with EcoRI and then separated on a 1% agarose gel. After capillary transfer, the membrane was probed with the 32P-labeled synthetic 17-mer probe. Lanes: 1, L. lactis subsp. lactis CC9; 2, L. lactis subsp. lactis CC9(pGKV259); 3 through 9, L. lactis subsp. lactis CC9(pSGK1.6R) isolates.

4 1112 KIM AND BATT newly procured resistances, in some cases as a result of a single-site mutation. A strain of L. lactis subsp. lactis has been engineered to be bacteriophage resistant by expressing the antisense strand of a conserved gene product. The ability of antisense mrna to inhibit gene expression has been demonstrated in a number of bacteria, plants, and cultured animal cells. The exact mechanism by which antisense mrna inhibits the expression of a targeted gene is not known. Most likely it acts at a posttranscriptional level by producing a nontranslatable double-stranded mrna hybrid and/or by destabilizing the sense mrna. Recent evidence in the micf-ompf system suggests that the formation of a hybrid mrna leads to its rapid degradation (1). In our study the resistance induced must be mediated by antisense mrna; there are no open reading frames of any significant size coded for on the antisense strand of the fragment. Furthermore, the construct with the sense strand positioned to be transcribed by the pgkv259- borne promoter does not affect bacteriophage infection. Preliminary experiments indicate that bacteriophage adsorption is not inhibited and that approximately 99% of the bacteriophage adsorb to the surface of L. lactis subsp. lactis CC9 whether or not it carries psgk1.6r. This suggests that the antisense mrna is preventing some later stage of bacteriophage replication or maturation. We have observed that only the entire 1,654-bp fragment in the antisense orientation is capable of inhibiting bacteriophage infection. In previous studies, a region coding for only the 5' end of the targeted gene but including the native ribosome binding site appeared to be sufficient for inhibition of a single-stranded RNA coliphage (3). A 247-bp fragment complementary to the ribosome binding site and translational start for the coat protein and a 159-bp fragment complementary to similarly located regions of the replicase protein resulted in an approximately 60% reduction in PFU. However, a 38% reduction in PFU was obtained with a 690-bp fragment that was complementary to the 3' end of the replicase gene. In our case, the antisense mrna construct psgk1.or, consisting of only the 695-bp EcoRI-EcoRV fragment, fails to inhibit bacteriophage replication, although it contains the putative translational start and ribosome binding site. It is unclear whether this truncated open reading frame fails to provide protection due to its inability to form a stable hybrid or whether a region other than or in addition to the 5' end must be targeted. Furthermore, the relative stability of each of the antisense mrna products may be a factor accounting for the observed differences (6). In preliminary Northern RNA blot analysis of L. lactis subsp. lactis CC9(pSGK1.6R) with an oligonucleotide primer as a probe for the mrna produced from the antisense strand, a prominent 3-kb transcript is observed. Although this provides partial proof for the production of an antisense mrna, conclusive evidence, including a quantification of the levels of both sense and antisense mrna from this conserved 1.6-kb region during infection, is required. An antisense mrna-mediated mechanism is suggested, since the inhibition is only observed when a promoter is positioned to transcribe the antisense strand. Antisense mrna has the unique attribute of potentially being immune to any single-site mutations in the bacteriophage genome that might circumvent other resistance mechanisms. Within the bacteriophage isolates tested, neither of two distinguishable groups, I and 11 (9), can infect L. lactis subsp. lactis CC9(pSGK1.6R). These groups can be classified on the basis of a polymorphic EcoRI restriction site and have at least two nucleotide sequence differences within the APPL. ENVIRON. MICROBIOL. region coding for the 51-kDa protein. Perfect homology between the antisense mrna and the targeted bacteriophage is not necessary for protection, since the two nucleotide sequence differences between these two groups of bacteriophage do not affect their ability to be inhibited. Therefore, these bacteriophages should not be able to infect L. lactis subsp. lactis CC9(pSGK1.6R) unless they are able to acquire a gene with a radically different coding sequence whose product can supplant the function of the 51-kDa protein. ACKNOWLEDGMENTS This work has been supported by New York State (NYC ), National Dairy Promotion and Research Board, USDA-CRGO (87-CRCR ), and a gift from Universal Foods, Milwaukee, Wis. We acknowledge the contributions of Jan Kok and discussions with David Wilson and Jeong Hwan Kim. We also thank Barbara Russell for her assistance in the preparation of the manuscript. REFERENCES 1. Andersen, J., S. A. Forst, K. Zhao, M. Inouye, and N. Delihas The function of micf RNA. J. Biol. Chem. 264: Anderson, D. G., and L. L. McKay Simple and rapid method for isolating large plasmid DNA from lactic streptococci. Appl. Environ. Microbiol. 46: Coleman, J., A. Hirasuma, Y. Inokuchi, P. J. Green, and M. Inouye A novel immune system against bacteriophage infection using complementary RNA (micrna). Nature (London) 315: Ecker, J., and R. Davis Inhibition of gene expression in plant cells by expression of antisense RNA. Proc. Natl. Acad. Sci. USA 83: Hanahan, D Studies on transformation of Escherichia coli with plasmids. J. Mol. Biol. 166: Hirashima, A., S. Sawaki, Y. Inokuchi, and M. Inouye Engineering of the mrna-interfering complementary RNA immune system against viral infection. Proc. Natl. Acad. Sci. USA 83: Izant, J., and H. Weintraub Inhibition of thymidine kinase gene expression by anti-sense RNA: a molecular approach to genetic analysis. Cell 36: Jarvis, A. W Bacteriophages of lactic acid bacteria. J. Dairy Sci. 72: Kim, S. G., and C. A. Batt Identification of a conserved nucleotide sequence in Lactococcus lactis subsp. lactis bacteriophages. Gene 98: Klaenhammer, T. R Plasmid-directed mechanisms for bacteriophage defense in lactic streptococci. FEMS Microbiol. Rev. 46: Kok, J., J. M. B. M. van der Vossen, and G. Venema Construction of plasmid cloning vectors for lactic streptococci which also replicate in Bacillus subtilis and Escherichia coli. Appl. Environ. Microbiol. 48: Maniatis, T., E. F. Fritsch, and J. Sambrook Molecular cloning: a laboratory manual. Cold Spring Harbor Press, Cold Spring Harbor, N.Y. 13. Marshall, R. J., and N. J. Berridge Selection and some properties of phage-resistant starter for cheese-making. J. Dairy Res. 43: McKay, L. L Functional properties of plasmids in the lactic acid bacteria streptococci. Antonie van Leeuwenhoek J. Microbiol. Serol. 49: Mizuno, T., M.-Y. Chou, and M. Inouye A unique mechanism regulating gene expression: translational inhibition by a complementary RNA transcript (micrna). Proc. Natl. Acad. Sci. USA 81: Pestka, S., B. L. Daugherty, V. Jung, K. Hotta, and R. K. Pestka Anti-mRNA: specific inhibition of translation of single mrna molecules. Proc. Natl. Acad. Sci. USA 81:7525-

5 VOL. 57, Stephenson, M. L., and P. C. Zamecnik Inhibition of Rous sarcoma viral RNA translation by a specific oligodeoxyribonucleotide. Proc. Natl. Acad. Sci. USA 75: Takayama, K. M., N. Houba-Herin, and M. Inouye Overproduction of an antisense RNA containing the oop RNA sequence of bacteriophage induces clear plaque formation. Mol. Gen. Genet. 210: Terzaghi, B. E., and W. E. Sandine Improved medium for lactic streptococci and their bacteriophages. Appl. Microbiol. 29: Thunell, R., W. Sandine, and F. Bodyfelt Phage-insensitive, multiple-strain starter approach to cheddar cheese. J. ANTISENSE mrna 1113 Dairy Sci. 64: van der Vossen, J. M. B. M., D. van der Lelie, and G. Venema Isolation and characterization of Streptococcus cremoris Wg2-specific promoters. Appl. Environ. Microbiol. 53: Vieira, J., and J. Messing The puc plasmids, an M13mp7-derived system for insertion mutagenesis and sequencing with synthetic universal primers. Gene 19: Whitehead, H. R., and G. A. Cox Bacteriophage phenomena in cultures of lactic streptococci. J. Dairy Res. 7: Yamamoto, K. R., and B. M. Alberts Rapid bacteriophage sedimentation in the presence of polyethylene glycol and its application to large-scale purification. Virology 40: